The NYT just had a very interesting article on risks associated with hedge fund. This piece, titled "Is a Hedge Fund Shakeout Coming soon? This Insider Thinks So" highlights some of the work done by Professor Andrew Lo of MIT. Professor Lo has done a lot of work with hedge funds and is involved on the practitioner side. His paper (available here) details some of the risks associated with these vehicles (caution: it's an academic paper and is written for "quant jocks", so its' pretty mathematically dense).
One of the key points of the NYT article is that hedge funds' returns are associated with risks that aren't nearly as common in other alternative investments. One of the major ones is "liquidity risk". A lot of the assets traded by hedge funds are infrequently traded. So, the reports issued by hedge funds profiles are either based on "stale" prices (if they don't trade, you don't have "market" prices), or are based on estimates of asset values made by the hedge fund sponsor. This makes allows the sponsor to make the hedge funds' reported returns appear much "smoother" than they actually are.
There are additional problem associated with liquidity risk. One of hedge funds' greatest appeals is that they are relatively uncorrelated with the market as a whole. However, this 'uncorelatedness" may not hold in times of market stress. In particular, their returns may become highly correlated in the event of an exodus of investor capital from these funds. This effect is driven by the same illiquidity of hedge fund assets - if you have to sell an illiquid asset, you have to accept a big discount to unload it quickly.
The article finishes by listing a number of factors that could trigger a hedge fund meltdown - like oil over $100 a barrel (I think the article was written before Katrina) or a tightening of rules at Fannie Mae.
All in all, a piece worth reading - click here for the whole thing.